[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] Virtual addressing



Guido Flohr wrote:
>Hi Michael,
>
>On Mon, Sep 06, 1999 at 01:40:09AM -0500, Michael White wrote:
>> Hi all,
>> 
>>    As some of you may (or may not) know, I've been working on virtual
>> addressing for MiNT.  This gives each process a potentially huge
>> address space, only limited by the amount of physical RAM in the
>> system.  This will remove the TT RAM fragmentation on 
>> machines with an MMU (i.e. 68030 and higher).  It will also be a step
>> toward having virtual memory.
>
>Sorry, the MMU is certainly not my area of expertise. ;-)
>This is just to tell you: Yes, I would be very interested in virtual
>memory.
>
>Besides, do you own a Milan?  If not, you should get in contact with
>Michael Schwingen and/or Frank Naumann.  AFAIK the Milan uses the MMU in
>its own way and that will lead to problems.  Don't know the details,
>sorry, but I know that there are things to take care of.

Hi Guido,

   No, I don't own a Milan, but I do own an Afterburner.  It too uses
the MMU in its own way.  I've got a dump of a Hades MMU tree set up
by ozk's programs.  I would like a Milan tree just to confirm what
I'm doing will be generic enough across all platforms.

   Note that I'm not doing virtual memory, I'm doing virtual
addressing.  The difference?  Virtual memory uses the hard disk to
make it look like your computer has gobs of memory.  Virtual 
addressing uses the MMU to make memory pages that may be located
anywhere look like contiguous memory to a process.  Also known as
paging.

    Virtual memory will give you lots of room to run your applications
(albeit a bit more slowly).  Virtual addressing will prevent memory 
fragmentation in physical (and virtual too, I suppose) RAM that appears
due to various processes starting and exiting.  This is usually not
seen unless your system is up for extended periods.


Michael White (michael@fastlane.net)